BitShares Forum

Main => General Discussion => Topic started by: Bitcoinfan on January 09, 2014, 01:17:22 am

Title: Coinjoin implementation
Post by: Bitcoinfan on January 09, 2014, 01:17:22 am
Are there plans to quickly implement Coinjoin at the protocol level in Bitshares?  If Invictus did this, it would be an enormous differentiating factor from all other alt-coins, if there wasn't enough already. It would make bitshares catch on like fire. 
Title: Re: Coinjoin implementation
Post by: ripplexiaoshan on January 09, 2014, 03:32:13 am
Has bitcoin wallet exploited this protocol? I searched in this post, but didn't get the answer, but I think the answer is no, right?  I will appreciate it if you can elaborate a little bit.   
https://bitcointalk.org/index.php?topic=279249.0
Title: Re: Coinjoin implementation
Post by: Troglodactyl on January 11, 2014, 01:25:49 am
Couldn't this be added into the Keyhotee integrated wallets rather than the protocol?

Title: Re: Coinjoin implementation
Post by: fuzzy on January 20, 2014, 08:49:53 pm
i'd be very interested in seeing this happen to.  Any responses forthcoming?
Title: Re: Coinjoin implementation
Post by: bytemaster on January 20, 2014, 09:54:56 pm
Coinjoin is an interesting concept, we are taking a slightly different approach:   Coin-Never-Join.   If 95% of all transfers have a single input and single output and no address is ever reused then you get the same effect.  After many such transactions performing a single join of two of your addresses to combine 'dust' doesn't actually convey much meaning that can be tracked.

Title: Re: Coinjoin implementation
Post by: vikram on January 21, 2014, 04:53:28 am
Coinjoin is an interesting concept, we are taking a slightly different approach:   Coin-Never-Join.   If 95% of all transfers have a single input and single output and no address is ever reused then you get the same effect.  After many such transactions performing a single join of two of your addresses to combine 'dust' doesn't actually convey much meaning that can be tracked.

Can you explain a bit more? What is different about BitShares X that allows people to use only single inputs/outputs and unique addresses 95% of the time whereas this isn't particularly common in Bitcoin?
Title: Re: Coinjoin implementation
Post by: bytemaster on January 21, 2014, 05:42:53 am
Coinjoin is an interesting concept, we are taking a slightly different approach:   Coin-Never-Join.   If 95% of all transfers have a single input and single output and no address is ever reused then you get the same effect.  After many such transactions performing a single join of two of your addresses to combine 'dust' doesn't actually convey much meaning that can be tracked.

Can you explain a bit more? What is different about BitShares X that allows people to use only single inputs/outputs and unique addresses 95% of the time whereas this isn't particularly common in Bitcoin?

It isn't BitShares X, it is the Keyhotee Wallet which has not been implemented yet but will allow this level of security with Bitcoin as well.   The key is Hierarchal wallets and multi-part 'transactions'.    Imagine giving someone a single "extended public key" that allowed them to generate as many addresses for you as they require.   Now when they want to send you 100 BTC, they can do so via 20 individual transactions (automatically) that your client can then recognize as being part of one 'logical' transaction.   



Title: Re: Coinjoin implementation
Post by: Bitcoinfan on January 21, 2014, 11:25:58 pm
Coinjoin is an interesting concept, we are taking a slightly different approach:   Coin-Never-Join.   If 95% of all transfers have a single input and single output and no address is ever reused then you get the same effect.  After many such transactions performing a single join of two of your addresses to combine 'dust' doesn't actually convey much meaning that can be tracked.

Can you explain a bit more? What is different about BitShares X that allows people to use only single inputs/outputs and unique addresses 95% of the time whereas this isn't particularly common in Bitcoin?

It isn't BitShares X, it is the Keyhotee Wallet which has not been implemented yet but will allow this level of security with Bitcoin as well.   The key is Hierarchal wallets and multi-part 'transactions'.    Imagine giving someone a single "extended public key" that allowed them to generate as many addresses for you as they require.   Now when they want to send you 100 BTC, they can do so via 20 individual transactions (automatically) that your client can then recognize as being part of one 'logical' transaction.

cool.  And it doesn't bloat the blockchain?  Im coming with the bitcoin perspective that more public keys, means bigger bitcoin dataset size. 
Title: Re: Coinjoin implementation
Post by: bytemaster on January 21, 2014, 11:59:23 pm
Transaction fees will be how the market balances security vs privacy.
Title: Re: Coinjoin implementation
Post by: Troglodactyl on January 22, 2014, 01:35:03 am
More methods are good.  It still seems as though there are significant advantages to the Coinjoin approach for certain scenarios, but if this is confirmed, someone will add it into the Keyhotee wallet later.